Predictive rightsizing for virtual machines in cloud computing systems

Abstract

Various embodiments, methods, and systems for implementing a predictive rightsizing system are provided. Predicted rightsized deployment configurations are generated for virtual machine “VM” deployments having deployment configurations that are modified to predicted rightsized deployment configurations based on a prediction engine. In operation, a VM deployment, associated with a request to deploy one or more VMs on a node, is accessed at a predictive rightsizing controller. A predicted resource utilization for the VM deployment is generated at the prediction engine and accessed at the predictive rightsizing controller. The predicted resource utilization is generated based on a prediction engine that uses past behaviors and features associated with previous VM deployments. Based on the predicted resource utilization, a predicted rightsized deployment configuration is generated for the VM deployment. The predicted rightsized deployment configuration is communicated such that the request to deploy the VMs is executed based on the predicted rightsized deployment configuration.

What is claimed is: 
     
       1. A predictive rightsizing system for rightsizing virtual machines, the system comprising:
 a processor and memory; and 
 machine readable instructions stored in the memory, when executed by the processor, configure the processor to: 
 access a virtual machine deployment associated with a request to deploy a virtual machine on a node from a plurality of nodes running a plurality of virtual machines in a cloud computing system; 
 access a predicted resource utilization for the virtual machine deployment, wherein the predicted resource utilization is generated based on a prediction engine that uses past behaviors and a plurality of features associated with previous virtual machine deployments to make resource utilization predictions, wherein predicted resource utilizations are associated with corresponding confidence scores that are used for selectively generating predicted rightsized deployments, wherein the predicted rightsized deployment configurations are generated when the confidence scores meet a confidence score threshold; 
 based on the predicted resource utilization, generate a predicted rightsized deployment configuration for the virtual machine deployment; and 
 communicate the predicted rightsized deployment configuration such that the request to deploy the virtual machine is executed based on the predicted rightsized deployment configuration. 
 
     
     
       2. The system of  claim 1 , wherein the predicted resource utilization is a predicted maximum resource utilization for the virtual machine deployment, wherein the predicted resource utilization is associated with one or more resource types, wherein the virtual machine deployment comprises a deployment configuration comprising the virtual machine and one or more resource types. 
     
     
       3. The system of  claim 1 , wherein the plurality of features are associated with previous virtual machine deployments in the cloud computing system; wherein, based on the plurality features, the past behaviors for the previous virtual machine deployments are used generate the predicted resource utilization. 
     
     
       4. The system of  claim 1 , wherein generating the predicted rightsized deployment configuration for the virtual machine deployment is based on the predicted resource utilization to circumvent a monitoring period for monitoring virtual machine deployments prior to generating a recommended rightsized deployment. 
     
     
       5. The system of  claim 1 , wherein generating the predicted rightsized deployment configuration further comprises: based on the predicted resource utilization, modifying one or more resources in a deployment configuration of the virtual machine deployment. 
     
     
       6. The system of  claim 1 , wherein the request is associated with a customer of the cloud computing system, wherein the customer has opted-in to automatically executing the request based on the predicted rightsized deployment configuration. 
     
     
       7. The system of  claim 1 , wherein the confidence scores and the confidence score thresholds are used to selectively deploy the predicted rightsized deployments, wherein the predicted rightsized deployments are not deployed when the confidence scores do not meet the confidence score thresholds. 
     
     
       8. One or more computer storage media having computer-executable instructions embodied thereon that, when executed, by one or more processors, cause the one or more processors to perform a method for rightsizing virtual machines, the method comprising:
 accessing a virtual machine deployment associated with a request to deploy a virtual machine on a node from a plurality of nodes running a plurality of virtual machines in a cloud computing system; 
 accessing a predicted resource utilization for the virtual machine deployment, wherein the predicted resource utilization is generated based on a prediction engine that uses past behaviors and a plurality of features associated with previous virtual machine deployments to make resource utilization predictions, wherein predicted resource utilizations are associated with corresponding confidence scores that are used for selectively generating predicted rightsized deployments, wherein the predicted rightsized deployment configurations are generated when the confidence scores meet a confidence score threshold; 
 based on the predicted resource utilization and a confidence score meeting the confidence score threshold, generating a predicted rightsized deployment configuration for the virtual machine deployment; and 
 communicating the predicted rightsized deployment configuration such that the request to deploy the virtual machine is executed based on the predicted rightsized deployment configuration. 
 
     
     
       9. The media of  claim 8 , wherein the predicted resource utilization is a predicted maximum resource utilization for the virtual machine deployment, wherein the predicted resource utilization is associated with one or more resource types, wherein the virtual machine deployment comprises a deployment configuration comprising the virtual machine and one or more resource types. 
     
     
       10. The media of  claim 8 , wherein the plurality of features are associated with previous virtual machine deployments in the cloud computing system, wherein, based on the plurality features, the past behaviors for the previous virtual machine deployments are used generate the predicted resource utilization. 
     
     
       11. The media of  claim 8 , wherein generating the predicted rightsized deployment configuration for the virtual machine deployment is based on the predicted resource utilization to circumvent a monitoring period for monitoring virtual machine deployments prior to generating a recommended rightsized deployment. 
     
     
       12. The media of  claim 8 , wherein generating the predicted rightsized deployment configuration further comprises: based on the predicted resource utilization modifying one or more resources in a deployment configuration of the virtual machine deployment. 
     
     
       13. The media of  claim 8 , wherein the request is associated with a customer of the cloud computing system, wherein the customer has opted-in to automatically executing the request based on the predicted rightsized deployment configuration. 
     
     
       14. The media of  claim 8 , wherein the confidence score threshold is selectively configured based a plurality of factors associated with the cloud computing system. 
     
     
       15. A method for rightsizing virtual machines, the method comprising:
 accessing a virtual machine deployment associated with a request to deploy a virtual machine on a node; 
 accessing a predicted resource utilization for the virtual machine deployment, wherein the predicted resource utilization is generated based on a prediction engine that uses past behaviors and a plurality of features associated with previous virtual machine deployments make resource utilization predictions, wherein predicted resource utilizations are associated with corresponding confidence scores that are used for selectively generating predicted rightsized deployments, wherein the predicted rightsized deployment configurations are generated when the confidence scores meet a confidence score threshold; 
 based on the predicted resource utilization, generating a predicted rightsized deployment configuration for the virtual machine deployment; and 
 processing the predicted rightsized deployment configuration. 
 
     
     
       16. The method of  claim 15 , wherein the predicted resource utilization is a predicted maximum resource utilization for the virtual machine deployment, wherein the predicted resource utilization is associated with one or more resource types, wherein the virtual machine deployment comprises a deployment configuration comprising the virtual machine and one or more resource types. 
     
     
       17. The method of  claim 15 , wherein the plurality of features are associated with previous virtual machine deployments in the cloud computing system; wherein, based on the plurality features, the past behaviors for the previous virtual machine deployments are used generate the predicted resource utilization. 
     
     
       18. The method of  claim 15 , wherein generating the predicted rightsized deployment configuration for the virtual machine deployment is based on the predicted resource utilization to circumvent a monitoring period for monitoring virtual machine deployments prior to generating a recommended rightsized deployment. 
     
     
       19. The method of  claim 15 , wherein the confidence scores and the confidence score thresholds are used to selectively deploy the predicted rightsized deployments, wherein the predicted rightsized deployments are not deployed when the confidence scores do not meet the confidence score thresholds. 
     
     
       20. The method of  claim 15 , wherein processing the predicted rightsized deployment configuration comprises performing one of the following:
 causing the request to deploy the virtual machine to be executed based on the predicted rightsized deployment configuration; and 
 communicating the predicted rightsized deployment configuration to an interface.